算法服务接口文档

API

• 接口：http://server-host/ner

• 方式：POST

• 参数：content

• 响应：

   {
       "code": 0,
       "data": {
   		"mood_type": "positive",
   		"mood_sub_result": "O",
   		"mood_sub_type": {
   			"A": 0.2,
   			"C": 0.3,
   			"E": 0,
   			"N": 0.1,
   			"O": 0.4
   		}
       },
       "message": "success"
   }	
  

  code: 0 -> success | 401 -> Unauthorized | 503 -> LSTM failed | 400 -> content required

算法接口

• function: Sentiment_lstm.lstm_predict

• arguments: string

• return:

   {
   	"mood_type": "positive",
   	"mood_sub_result": "O",
   	"mood_sub_type": {
   		"A": 0.2,
   		"C": 0.3,
   		"E": 0,
   		"N": 0.1,
   		"O": 0.4
   	}
   }
  

---

问题

• flask 环境下调用算法接口，第二次之后报错：

   Cannot interpret feed_dict key as Tensor: Tensor Tensor("Placeholder_2:0", shape=(50, 200), dtype=float32) is not an element of this graph.
  
   解决办法：
   keras.backend.clear_session()
  

• nginx 502 --> 代码 ImportError(ImportError: libcudnn.so.5: cannot open shared object file: No such file or directory) --> cuda 库软的链接错误（symbol link）。解决方法：

   # 查看 cudnn 库相关的信息，找到错误的链接
   $ sudo ldconfig -v | grep "libcudnn"
   /sbin/ldconfig.real: /usr/local/cuda-8.0/targets/x86_64-linux/lib/libcudnn.so.5 is not a symbolic link
  
   libcudnn.so.5 -> libcudnn.so.5.1.10
  
   # 重新建立链接
   sudo ln -sf /usr/local/cuda-8.0/targets/x86_64-linux/lib/libcudnn.so.5.1.10 /usr/local/cuda-8.0/targets/x86_64-linux/lib/libcudnn.so.5